Subscribe RSS

Archive for the ‘Research’ Category

Similarité ou ressemblance?

March 5th, 2010 by fmn | No Comments | Filed in Research

Translate original post with Google Translate

L’asymétrie est une propriété essentielle du jugement humain de la similarité entre deux concepts ou stimuli, comme l’a illustré et modélisé Amos Tversky 1. Dans un exemple, maintenant classique, Tversky compare politiquement la Corée du Nord et la Chine. Deux propositions sont possibles :

(Exemple 1)

(a) La Chine est similaire à la Corée du Nord.
(b) La Corée du Nord est similaire à la Chine.

Ces deux propositions ne sont pas équivalentes, la même importance ne leur est pas attachée. Cette asymétrie est observée pour nombreux exemples, y compris pour des stimuli visuels 2.

Si l’existence de cette asymétrie n’est pas contestable, il n’est pas forcément immédiat de la sentir dans les propositions (a) et (b). Mon avis est que le terme similarité ne permet pas, en français, de bien sentir cette directivité. Je propose d’utiliser à la place le terme ressemblance (ou le verbe ressembler):

(Exemple 2)

(a) La Chine ressemble à la Corée du Nord.
(b) La Corée du Nord ressemble à la Chine.

Je trouve ainsi beaucoup plus clair dans l’exemple (2) que la proposition (b) semble plus conforme à la réalité que la proposition (a). La similarité de la Chine à la Corée du Nord est plus grande que la similarité de la Corée du Nord à la Chine.

L’interprétation de Tversky est que la ressemblance (similarity) entre la variante et le prototype est plus faible qu’entre le prototype à la variante. Je ferai de cette interprétation l’objet de mon prochain billet.

FMN.

1.

  • [1977,article] bibtex
    A. Tversky, "Features of similarity," Psychological Review, vol. 84, pp. 327-352, 1977.
    @article{Tversky77,
      author = {A. Tversky},
      title = {Features of similarity},
      journal = {Psychological Review},
      volume = {84},
      pages = {327--352},
      year = {1977}
    }

2.

  • [1975,article] bibtex Go to document
    E. Rosch, "Cognitive reference points," Cognitive Psychology, vol. 7, iss. 4, pp. 532-547, 1975.
    @article{Rosch75,
      author = {Rosch, Eleanor},
      citeulike-article-id = {1116906},
      citeulike-linkout-0 = {http://dx.doi.org/10.1016/0010-0285(75)90021-3},
      doi = {10.1016/0010-0285(75)90021-3},
      journal = {Cognitive Psychology},
      keywords = {\_d\_cognitive-biases, \_d\_deep-concepts, category-learning-use, empirical},
      month = {October},
      number = {4},
      pages = {532--547},
      posted-at = {2007-02-21 19:54:24},
      priority = {2},
      title = {Cognitive reference points},
      url = {http://dx.doi.org/10.1016/0010-0285(75)90021-3},
      volume = {7},
      year = {1975}
    }

Tags: ,

A minimal ImageJ Plugin in Clojure: image inversion

January 15th, 2010 by fmn | 2 Comments | Filed in Enseignement, Research

I show in this post how to write an ImageJ plugin with Clojure. This example is taken from Digital Image Processing: An Algorithmic Introduction Using Java: an image inversion (page 32).

The goal is to invert all the pixels of a 8-bit grayscale image, turning an image into its negative. As a pixel value is coded with 8 bits, the higher possible value is 255. The operation is thus to transform each pixel value v into 255-v.

I first present the plugin in Java, with a description of the essentials elements of an ImageJ plugin. Then, i give several Clojure versions. The last is as fast as the Java one, but more reusable.

(more…)

Tags: , , , , , , , ,

A new open access journal : IPSJ Transactions on Computer Vision and Applications

January 4th, 2010 by fmn | No Comments | Filed in Research

This year begins with good news, as a new journal is announced :  IPSJ Transactions on Computer Vision and Applications.

The covered areas are fundamental and applied computer vision. So Classical. But the special feature of the journal is to be on line and open access. Thus every article is available and can be read, downloaded and printed free at no cost. If the point for the reader is obvious, for the editors the objective is to obtain high exposure. I wish them great success (this is why i make some advertising).

In the first issue, i have noted and placed in my lecture list the following articles:

  • Content-based Image Retrieval by Indexing Random Subwindows with Randomized Trees (Raphaël Marée, Pierre Geurts and Louis Wehenkel)
  • A Survey of Manifold Learning for Images (Robert Pless and Richard Souvenir)
  • Partial Similarity of Shapes Using a Statistical Significance Measure (Alexander M. Bronstein, Michael M. Bronstein, Yair Carmon and Ron Kimmel)
  • and Using Context to Recognize People in Consumer Images (Andrew C. Gallagher and Tsuhan Chen)

A quite good selection for this first issue. Nice !

FMN.

Tags: , ,